Tr a n s l a t i o n a l O n c o l o g y
Volume 8 Number 4
August 2015
pp. 231–238 231
www.transonc.com
Dasatinib Modulates Invasive
and Migratory Properties of
Canine Osteosarcoma and
has Therapeutic Potential in
Affected Dogs
Kevin Marley, Justine Gullaba 1, Bernard Seguin 2,
Howard B. Gelberg and Stuart C. Helfand
Department of Clinical Sciences Oregon State University,
105 Magruder Hall, Corvallis OR, 97331, USA
Abstract
BACKGROUND: This investigation sought to elucidate the relationship between hepatocyte growth factor (HGF)–
induced metastatic behavior and the tyrosine kinase inhibitors (TKIs) crizotinib and dasatinib in canine
osteosarcoma (OS). Preliminary evidence of an apparent clinical benefit from adjuvant therapy with dasatinib in
four dogs is described. METHODS: The inhibitors were assessed for their ability to block phosphorylation of MET;
reduce HGF-induced production of matrix metalloproteinase (MMP); and prevent invasion, migration, and cell
viability in canine OS cell lines. Oral dasatinib (0.75 mg/kg) was tested as an adjuvant therapy in four dogs with OS.
RESULTS: Constitutive phosphorylation of MET was detected in two cell lines, and this was unaffected by 20-nM
incubation with either dasatinib or crizotinib. Incubation of cell lines with HGF (MET ligand) increased cell migration
and invasion in both cell lines and increased MMP-9 activity in one. Dasatinib suppressed OS cell viability and HGFinduced invasion and migration, whereas crizotinib reduced migration and MMP-9 production but did not inhibit
invasion or viability. CONCLUSIONS: Invasion, migration, and viability of canine OS cell lines are increased by
exogenous HGF. HGF induces secretion of different forms of MMP in different cell lines. The HGF-driven increase
in viability and metastatic behaviors we observed are more uniformly inhibited by dasatinib. These observations
suggest a potential clinical benefit of adjuvant dasatinib treatment for dogs with OS.
Translational Oncology (2015) 8, 231–238
Introduction
The receptor tyrosine kinase (RTK) MET is considered a growth and
motility factor because it provides crucial signals for survival and
migration during embryogenesis [1]. MET signaling is normally
initiated by binding its ligand, hepatocyte growth factor (HGF, also
known as scatter factor), but may be constitutively activated as a
consequence of point mutations, coexpression with HGF in the same
tissue, or overexpression [2]. Deregulation of MET is associated with
transformation and metastatic progression in many cancers, and this has
led to research of the MET pathway as a therapeutic target [3,4].
Anti-MET strategies have been investigated using antibodies [5], decoy
receptors [6], and a variety of tyrosine kinase inhibitors (TKIs) [7–10].
MET is expressed in a variety of tissues including canine
osteosarcoma (OS) tumors and cell lines [11]. The MET-specific
TKI crizotinib, when tested against human OS cell lines, was able to
selectively induce apoptosis, inhibit cell growth in vitro, and prevent
tumor growth in a mouse xenograft model [10]. The overexpression
and coexpression of MET and HGF in canine OS suggest autocrine
or paracrine activation and indicate that MET may be an important
target in this neoplasm [12,13]. In this regard, short interfering RNA
against MET reduced motility and invasion in vitro[13]; but the
MET-specific TKI, PHA-665752, had no effect on viability in canine
OS cells expressing high levels of MET [13]. This suggests the MET
Address all correspondence to: Stuart C. Helfand or Kevin Marley, Department of Clinical
Sciences Oregon State University, 105 Magruder Hall, Corvallis Oregon, 97331, USA.
E-mail: sch_726@yahoo.com (Stuart C. Helfand), kevin.marley@oregonstate.edu
(Kevin Marley)
1
Present address: Purdue University College of Veterinary Medicine, 625 Harrison St,
West Lafayette, IN, 47907, USA.
2
Present address: Flint Animal Cancer Center, Department of Clinical Sciences
Colorado State University, Fort Collins, Colorado, 80525, USA.
Received 6 January 2015; Revised 18 March 2015; Accepted 24 March 2015
© 2015 The Authors. Published by Elsevier Inc. on behalf of Neoplasia Press, Inc.
This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).
1936-5233/15
http://dx.doi.org/10.1016/j.tranon.2015.03.006
232
Dasatinib Migration and Invasion in Canine Osteosarcoma
Marley et al.
pathway, in some cases, may act independently from those that
modulate cell survival; and this could be important to the study of
metastatic behavior in tumor cells.
Metastatic tumor cells may have hijacked the MET pathway to
increase their ability to invade surrounding tissue [14]. This
complicated process likely involves coordinated changes in cell-to-cell
signaling, the induction of contact independence, and the production of
proteases, such as matrix metalloproteinase, that degrade extracellular
matrix [15]. In the current study, the MET pathway was investigated in
canine OS in the context of invasive behavior, including matrix
metalloproteinase (MMP) activity induced by exogenous HGF. The
TKIs crizotinib and dasatinib were tested for their ability to reduce cell
viability, prevent invasion and migration, and block MMP activity in
the presence and absence of HGF. In addition, our previous work
identified dasatinib as a potential treatment for canine OS [16].
Although MET has not been identified as a specific target of dasatinib,
dasatinib has been reported to block invasion, migration, and
proliferation in a variety of human sarcoma cell lines including OS
[17]. Herein we report the apparent extended survival of a cohort of four
dogs with OS that were treated with dasatinib.
Methods
Cell Lines and Reagents
Two canine OS cell lines, COS [18] and Clone-4 developed in our
laboratory (BS), were used in this study. Cells were cultured at 37°C
in a humidified 5% CO2 atmosphere. Cell culture medium was
RPMI 1640 supplemented with 2 mM glutamine, 2 mM sodium
pyruvate, 2 mM HEPES, 1% penicillin-streptomycin, and 10% FBS,
hereafter referred to as R10. Serum-free medium referred to as R0 was
identical except that it contained no FBS. Recombinant baculovirus
canine HGF (rcHGF) was purchased from R&D Systems (Minneapolis,
MN). Cells were detached with 0.25% trypsin and 0.03% EDTA
solution, and counted on a hemocytometer to assess cell numbers
before use.
Invasion Assay
Invasion studies were performed using 8-μm pore size, reduced
growth factor, Matrigel Invasion Chambers (BD Biosciences, Bedford,
MA) according to the manufacturer’s instructions. Briefly, cells were
seeded at a density of 25,000/well into the top chamber in R0 medium.
The bottom chambers contained R10 medium. HGF or TKI was added
to both the top and bottom chambers when appropriate, so the only
difference between the top and bottom chambers was the 10% serum.
After 24 hours, the membranes were removed, stained with Diff Quik,
mounted on glass slides, and imaged on a microscope. All cells that had
migrated through each membrane were counted, and each experiment
was repeated at least twice.
migrated into the previously clear zones were counted manually. To
accomplish this task, identically sized boxes were overlaid onto each
image using photo processing software (Adobe Photoshop); and all
cells within these were counted. The crosshatch pattern was used to
place the counting box in a similar location for every condition. Each
of these assays was repeated at least twice, and data represent the mean
of two independent assays.
Vital Staining
Cells prepared as described for migration assays (above) were
incubated with 0 to 40 nM dasatinib for 24 hours and then
trypsinized, exposed to trypan blue, and counted in a hemocytometer.
Cells that showed a deep blue color were considered dead. The
incubation medium was added back to the trypsinized cells before
counting to account for floating cells. Data shown are means ± SD of
two independent experiments.
Cell Viability
Cells were seeded in 96-well plates at a density of 5000 cells/well in a
volume of 100 μL of R10 and allowed to adhere overnight.
The following day, the R10 medium was replaced with R0; and the
cells were again incubated overnight. The medium was then
replaced with fresh R0 medium containing various concentrations of
TKI plus or minus 20 ng/mL of HGF. Cell viability was assessed after 48
hours using an 3-(4,5-dimethylthiazol-2-yl)-5-(3carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS)
assay according to the manufacturer’s instructions (Promega,
#G3580). All data are expressed as means ± SD of triplicate wells and
are representative of three independent experiments.
Flow Cytometry
Flow cytometry was used to detect binding of annexin V to
phosphotidyl serine on the surface of cells in early apoptosis and
propidium iodide to DNA in late-stage apoptosis or dead cells. Cells were
serum-starved overnight in R0 medium then incubated in the same
medium with 0 or 20 nM dasatinib ± HGF (20 ng/mL) for 24 hours. After
incubation, the media were collected and the cells collected using 0.005%
trypsin in PBS. The incubation media containing cells that were released
from the flask were then added back to the trypsinized cells along with a
100-μL aliquot of serum to stop the trypsin. Cells were then reacted with
biotinylated annexin V-FITC and propidium iodide using a commercially
available kit (Calbiochem). A minimum of 20,000 cells per sample was
collected on a Beckman Coulter FC-500 flow cytometer. Data analysis
and software compensation were performed using WinList (Verity
Software, Topsham, ME). Annexin V (FITC) and propidium iodide
fluorescence intensity were gated on forward and side scatter to exclude cell
debris. Spectral overlap was compensated using single-stain controls.
Western Blots
Migration Assay
5
Cells were seeded into six-well plates (2 × 10 cells/well) in 2 mL of
R10 medium and allowed to adhere overnight. The following day, the
R10 medium was replaced with R0; and the cells were again
incubated overnight. After 24 hours, the medium was removed and
replaced as described in each figure; and a sterile pipette tip was used
to scratch two intersecting lines in a cross configuration into the
subconfluent cells. The plates were imaged immediately using a
microscope with attached digital camera (Nikon Eclipse ti). After 20
hours, the scratched sections were imaged again; and cells that had
Translational Oncology Vol. 8, No. 4, 2015
Cells were seeded (3 × 10 6/well) in 25-cm 2 flasks and allowed to
adhere overnight before incubation with dasatinib or crizotinib in the
presence or absence of HGF. At the end of the incubation period, the
media and cells were removed using a cell scraper; and the cells were
pelleted in a tabletop centrifuge (3 minutes, 450g). The supernatants
were discarded, and cell pellets were extracted in 50-μL ice-cold
radioimmunoprecipitation buffer with added protease and phosphatase inhibitor cocktail (Thermo Scientific, Rockford, IL). Extracts
were sonicated 4 × 1 second using an ultrasonic dismembranator
(Fisher, Model 150T) and pelleted at 10,000g to remove cellular
5
HGF 24 hrs
ot
in
ib
D
+
F
+
G
F
HGF + crizotinib
24 hrs
HGF + dasatinib
24 hrs
Clone-4
COS
Control time 0
H
H
G
G
F
H
B
as
at
in
ib
0
+
H
G
F
C
riz
ot
in
ib
F
+
D
as
at
in
ib
0
10
C
riz
20
*
H
G
40
233
15
H
G
F
*
*
60
Marley et al.
Clone-4
20
R
0
*
Number of migrated cells
Dasatinib Migration and Invasion in Canine Osteosarcoma
COS
80
R
0
A
Number of migrated cells
Translational Oncology Vol. 8, No. 4, 2015
Figure 1. Migration assay. (A) HGF increased migration in both cell lines, but the TKIs crizotinib and dasatinib were able to suppress migration
only in the COS cell line. *P ≤ .05. Data are representative of three independent experiments. (B) Representative photomicrographs of cells
treated with HGF ± TKIs. After 24 hours, cells that migrated into the box overlay were counted (data shown in A).
debris. Protein concentration was measured using a Bradford assay
(Pierce, Rockford, IL). Proteins (20 μg/lane) were separated on 4% to
12% SDS polyacrylamide gels and transferred to polyvinylidene
COS
Clone-4
Percent Control
400
300
200
difluoride membranes using standard methods as follows. The
membranes were blocked in 2.5% albumin and reacted with
indicated primary antibodies diluted 500:1 for 3 hours at room
temperature or overnight at 4°C. The phospho-Met (Y1230/Y1234/
Y1235) antibody used was a rabbit polyclonal purchased from
Millipore (Billerica, MA). Total MET and actin antibodies were
purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The
membranes were washed, reacted with horseradish peroxidase–linked
secondary antibody (Santa Cruz Biotechnology) diluted 30,000:1,
and exposed to substrate (GE Healthcare).
Zymography
* *
100
F
G
H
H
G
F
+
+
C
D
riz
as
a
ot
tin
in
ib
ib
F
H
G
C
on
tr
o
l
0
Figure 2. Invasion assay. HGF increased invasion in both cell lines,
and this increase was prevented by dasatinib but not crizotinib. *P ≤ .05,
different from HGF treated. Data are presented as means ± SE of
duplicate assays and are representative of three independent assays.
Cells were seeded at a density of 3000 cells/well in 24-well plates
and allowed to adhere overnight. The following day, the growing cells
were serum starved for 24 hours and then incubated for 24 hours in
500 μL of R0 medium plus or minus either TKI at a concentration of
20 nM plus or minus HGF (20 ng/mL). After incubation, an aliquot
of medium was removed and centrifuged at 10,000g for 10 minutes
to remove debris. A 10-μL aliquot of each of these was separated by
electrophoresis on 10% polyacrylamide gelatin zymogram gels
according to the manufacturer’s directions (Invitrogen, Carlsbad,
CA). Following electrophoresis, the gels were incubated in developing
solution for 20 hours at 37°C, stained with Coomassie Blue
(G-BioSciences, St. Louis, MO), and imaged with a digital camera
(LAS 4000; GE Healthcare, Buckinghamshire, UK). Densitometry
234
Dasatinib Migration and Invasion in Canine Osteosarcoma
Marley et al.
Translational Oncology Vol. 8, No. 4, 2015
COS
*
**
*
100
125
Percent Control
Percent Control
125
Clone-4
Cntrl
HGF
*
75
50
25
*
* **
100
*
75
50
25
0
0
-2
-3
-1
0
1
2
3
4
-3
Log [dasatinib] (nM)
-2
-1
0
1
2
3
4
Log [dasatinib] (nM)
COS
Cntrl
HGF
Clone-4
Cntrl
HGF
125
100
125
Percent Control
Percent Control
Cntrl
HGF
75
50
25
100
75
50
25
0
0
-2
-3
-1
0
2
1
3
-3
4
-2
-1
0
1
2
3
4
Log [crizotinib] (nM)
Log [crizotinib] (nM)
Figure 3. MTS assays with tyrosine kinase inhibitors and HGF. Viability of HGF-stimulated cells was inhibited at low concentrations by
dasatinib (top panels) but not by crizotinib (bottom panels). *Difference between HGF-treated and control at the same TKI concentration
(P ≤ .05). IC50 values for dasatinib were 0.1 nM in controls and 32 nM in cells incubated with HGF in the COS cell line. These values were
0.3 nM and 112 nM in the Clone-4 cell line, respectively. IC50 values for crizotinib were 340 nM and 357 for COS controls and
HGF-treated cells and was not reached in the Clone-4 cell line.
for these and the Western blot images was performed using Image
Quant TL software (GE Healthcare).
Histopathology
Biopsy samples of the proximal tibia were decalcified, sectioned, and
stained with hematoxylin and eosin using standard methods.
B
5
10
20
10
0
40
te
d
da
da
s
at
sa
in
ti
ib
C
lo nib
ne
+
H
-4
un GF
C
tr
lo
C
ea
ne
lo
te
ne
-4
d
-4
da
da
sa
sa
tin
tin
ib
ib
+
H
G
F
2.5
20
tr
ea
0
Early Apoptotic
Late Apoptotic/necrotic
C
O
S
C
O
S
un
[dasatinib] (nM)
S
Clone-4
30
O
COS
C
45
40
35
30
25
20
15
10
5
0
Percent total gated events
Percent total
A 50
Statistics
Statistical comparisons using analysis of variance with Tukey post hoc
tests were performed, when appropriate, using Graphpad Prism
software (La Jolla, CA). Nonlinear regression using log-transformed
values were used to calculate IC50s from the MTS assay results using
the same software.
Figure 4. Cell death/apoptosis. (A) Trypan blue, live/dead, cell assay of COS and Clone-4 cells incubated 24 hours in 0- to 40-nM
concentrations of dasatinib shows no significant increase in cell death due to TKI treatment (P N .05). Data shown are means ± SD of two
independent experiments. (B) Flow cytometry data from cells incubated for 24 hours with 20 nM dasatinib show no increase in combined
level of early apoptotic or late apoptotic/necrotic cells (P N .05). Data shown are means ± SD of two independent experiments.
Translational Oncology Vol. 8, No. 4, 2015
Dasatinib Migration and Invasion in Canine Osteosarcoma
Marley et al.
235
A
A
Dasatinib Crizotinib HGF -
+
+
+
+
+
-
+
+
+
+
+
pY Met
Total Met
Pro MMP2 __
MMP2 __
Actin
COS
Clone-4
R0
HGF
HGF + Criz
HGF + Das
1.5
1.0
0.5
Clone-4
B
160
140
120
100
80
60
40
20
0
COS
Clone-4
-4
ne
C
lo
O
S
0.0
C
Normalized Density pY-MET
(arbitray units)
B
MMP2 activity percent
of HGF-only treated
condition
COS
Figure 5. MET activation. (A) Western blots from serum-starved cells
show that MET is constitutively activated in both cell lines. HGF
incubation had little effect on MET phosphorylation at tyrosines
1230/1234/1235, and TKI incubation (20 nM) was not sufficient to
reverse this. Dasatinib suppressed but did not eliminate the pY-MET
signal in Clone-4 cells. (B) Histogram of densitometry of pY-MET
Western blot shown in A.
Figure 6. Expression and protease activity of MMP2 in COS and
Clone-4 cells exposed to HGF and crizotinib or dasatinib. (A) Cells
were incubated overnight in R10 media (10% serum) or serum-free
media with HGF ± TKI as indicated. Criz-20 and criz-40 indicate
20-nM and 40-nM incubation concentrations. The expression
patterns between the two cell lines are quite different, with COS
cells expressing the pro- and active form uniformly, whereas the
Clone-4 cells expressed only the active form. The protease activity
of active MMP2 was reduced by both crizotinib and dasatinib, with
dasatinib appearing to be the more effective drug at these low
concentrations.
Results
Migration/Invasion
HGF caused an increase in migration in serum-starved canine OS
cells (P ≤ .05, Figure 1), and low-concentration (20 nM) TKI
incubation suppressed migration in COS but not Clone-4 cells. The
increase in migration attributed to HGF was significant in both cell
lines, but the rate of migration of the Clone-4 cells was much lower
(Figure 1). HGF-induced invasion was more consistent and was more
than 100% greater than controls in both cell lines (P b .05, Figure 2).
The TKIs' ability to prevent this increase was specific only to
dasatinib. In fact, dasatinib not only prevented the HGF-induced
increase but reduced overall invasion from 205 to 25% of that seen in
the serum-starved and untreated cells (Figure 2).
Cell Viability
Low-concentration dasatinib, but not crizotinib, reduced the
viability of serum-starved OS cells (Figure 3). This effect was blocked
by HGF in both cell lines at TKI concentrations less than 100 nM.
The protective effect of HGF was lost at dasatinib concentrations
greater than 100 nM. Crizotinib was less effective and only slightly
inhibited viability of the COS cells at concentrations of 100 nM and
greater (Figure 3). Crizotinib had no effect on the Clone-4 cell line at
concentrations less than 3 μM, which was the highest concentration
deemed reasonable in this study. No significant changes in cell death
or apoptosis were attributable to 20-nM dasatinib incubation
(Figure 4), which was the concentration used for the migration and
invasion assays. Under serum-starved conditions, the dasatinib IC50
values for the COS cell line were 0.1 nM for the untreated and 32 nM
for the HGF-treated cells. Dasatinib was similarly effective in the
Clone-4 cell line, with IC50s of 0.3 nM and 112 nM in the untreated
and treated cells, respectively. The crizotinib IC50s for the COS cells
were 340 nM and 357 nM for the untreated and HGF-treated cells,
respectively; and no IC50 values were observed for crizotinib in the
Clone-4 cell because the drug did not effectively reduce viability in
these cells.
MET Phosphorylation
Tyrosine phosphorylation (pY) of MET in serum-starved cells was
strongly positive and appeared unchanged by HGF (Figure 5). TKI
incubation similarly appeared to have little effect on pY-MET with
the exception that dasatinib was able to reduce, but not eliminate, this
activation signal in the Clone-4 cells.
Effects of HGF and TKIs on MMP-2 and MMP-9
This study assessed the effects of HGF on MMP-2 and MMP-9
secretion by OS cell lines and measured the ability of the TKIs
dasatinib and crizotinib to block this secretion. We were surprised by
the observation that MMP isoform production was dramatically
different in the two OS cell lines. Beginning with MMP-2, secretion
of the pro-form of MMP-2 was dramatically induced in the COS cells
by serum-containing media; but this was not observed in the Clone-4
236
Dasatinib Migration and Invasion in Canine Osteosarcoma
Marley et al.
A
__
__
48 kD
__ MMP-9
B
MMP-9 activity (fold
change normalized to
untreated)
100 kD
25
20
15
10
5
0
R0
HGF
HGF + Criz HGF + Das
Figure 7. Expression and protease activity of MMP9 in COS cells
exposed to HGF and crizotinib or dasatinib. (A) Zymogram showing
the result of two independent experiments ran on the same gel. (B)
Densitometry histogram of data from zymogram shown in A. Error
bars indicate SD.
cells (Figure 6). Both the pro- and active forms of MMP-2 were
produced in the presence of HGF in the COS cells, but HGF induced
only the active form in the Clone-4 cells. The zymogram
densitometry shows a dose–response relationship, with both TKIs
suppressing secretion of the active form of MMP-2; however, this
effect was more pronounced with dasatinib (Figure 6). The difference
between the two cell lines was more pronounced with MMP-9
secretion versus that seen in MMP-2, as media from the COS cells
produced a major band on the zymographs, whereas none was
detected using media from the Clone-4 cell line. In this regard, the
MMP-9 secretion was entirely dependent on HGF in the COS cells
(Figure 7, MMP-9 data for Clone-4 are not shown). This
HGF-induced MMP-9 secretion was suppressed in COS cells by
coincubation with crizotinib but not dasatinib. Data shown are
representative of three independent experiments (Figure 7).
Clinical Correlates
We have treated dogs with dasatinib for appendicular OS with
owner consent. This was combined with limb amputation and
carboplatin chemotherapy, which are among the current standardof-care treatments for canine OS. Although not a clinical trial, we felt
that it was important to determine if our in vitro data had the
potential to be translated into affected dogs in a pilot data setting.
This afforded the opportunity to gain experience on tolerable drug
dosing, establish a protocol for a future clinical trial, and begin to
generate observations of a potential therapeutic role for dasatinib in
dogs with OS.
The starting dose of dasatinib in all dogs has been 0.5 mg/kg given
every day or every other day (Table 1). Oral antiemetics were given on
treatment days if needed (maropitant 2 mg/kg once and/or ondansetron
0.2 mg/kg two to three times a day). The primary adverse effects have
Translational Oncology Vol. 8, No. 4, 2015
been inappetence and gastrointestinal upset with diarrhea. In most cases,
this has not precluded dasatinib treatment. After 2 weeks
without significant adverse effects, the 0.5-mg/kg dose was increased to
0.75 mg/kg and continued indefinitely or until the owner’s decision to
stop. Doses higher than 0.75 mg/kg were generally not tolerated. Table 1
contains details of the dogs’ signalment, tumor location, treatments, and
survival times. Two dogs’ tumors had been profiled against a panel of
drugs, and dasatinib was predicted to be the most effective in these
animals. These dogs had the longest survival times (Table 1). Details have
been previously published for the golden retriever in the table [16]. The
dogs in this report were all of large stature, which is typical of canine and
human OS patients. All developed OS in the proximal tibia, a common
site for OS in humans, whereas the distal radius is the most common site
in the dog. Histopathological examination by a board-certified veterinary
pathologist confirmed the diagnosis of OS in each dog’s tumor.
Representative histological sections from the tibial OS of the Great
Pyrenees (Table 1) are shown in Figure 8(A and B). The radiographic
image of this tumor is shown in Figure 8(C). It is primarily an osteolytic
tumor with destruction of bone within the medullary space and a
pathological fracture of the cortex (Figure 8C, lower arrow).
Three of the four dogs commenced dasatinib treatment after limb
amputation and five cycles of carboplatin chemotherapy given every 3
weeks at standard dosing (300 mg/m 2). Dasatinib treatment was
begun 3 weeks after completion of chemotherapy. The dogs were free
of radiographically detectable pulmonary metastases at that time. The
longest surviving dog (German Shepherd mix, Table 1) started
dasatinib when pulmonary metastases were detected after three of five
planned carboplatin treatments. A single (1 cm) pulmonary nodule
(Figure 8D) appeared to resolve after 3 months dasatinib treatment
(Figure 8E). This dog did not receive any additional carboplatin
chemotherapy but continued on dasatinib for a total of 25 months.
Thus, our preliminary radiographic evidence suggests that dasatinib
stabilized or induced partial remission of pulmonary metastases in at
least two dogs.
The shortest survival time of any of the four dogs treated with
adjuvant dasatinib was nearly 1.5 times as long as dogs noted in the
historical reports; and one dog, as of this writing, continues to be alive
33 months after diagnosis (N 3.2 times longer than the reported
median survival time).
Discussion
This study investigated the role of HGF in canine OS using an
in vitro model of metastasis because metastasis, not the primary
tumor, is the cause of mortality in OS. We describe effects of the
TKIs crizotinib and dasatinib using this model system and include
Table 1. Patient Characteristics, Survival and Treatments
Dog Signalment
Tumor Location
Survival Time
Dasatinib Treatment
Golden Retriever
7 y old, MN
Labrador Retriever
4 y old, MN
German Shepherd mix
9 y old, MN
Great Pyrenees
10 y old, MN
Left proximal tibia
29 mo *
Left proximal tibia
28 mo *
Right proximal tibia
33 mo; still living
Right proximal tibia
15 mo *
0.5-0.75 mg/kg
6.5 mo
0.5-0.75 mg/kg
10 mo
0.5-0.75 mg/kg
25 mo
0.5-0.75 mg/kg
7 mo
Q24, every 24 hours; EOD, every other day; MN, male neutered.
* Death due to metastatic disease.
Q24
EOD
Q24
EOD
Translational Oncology Vol. 8, No. 4, 2015
Dasatinib Migration and Invasion in Canine Osteosarcoma
Marley et al.
237
Figure 8. Case studies. (A) Histopathology of right tibial OS from the Great Pyrenees shows that the medullary cavity was invaded by
streams of tightly-packed, infiltrating, spindloid cells. These cells had fibrillar, eosinophilic, indistinctly-bordered cytoplasm and elongate
nuclei with undulant membranes, coarse, dispersed chromatin, and multiple nucleoli. (B) The cells became more rounded (arrow) along
seams of osteoid (asterisk) where they were trapped in lacunae. Multinucleated cells were present and there was 1 mitotic figure in 10
contiguous 400 × fields. (A) 200 ×, (B) 400 ×. (C) Radiographic image (lateral view) of the Great Pyrenees’ tibia shows marked destruction
of bone within the medullary space giving the appearance of patchy bone loss (top arrow). There is a pathological fracture of the cortex
(lower arrow). (D) Radiographic image of lung metastasis that appeared resolved after 3 months of dasatinib treatment (E). This dog
remains alive 33 months after diagnosis.
treatment and survival data from four dogs given dasatinib as an
adjuvant treatment for OS.
The study of metastatic behavior is complicated by the fact that
some tumor cells may harbor one or more mutations that drive both
metastasis and antiapoptotic or proliferative mechanisms. In this
study, we used low-concentration TKI incubations ± HGF as a way to
observe the HGF-specific behavior of cells in the presence of sublethal
concentrations of TKI. Higher TKI concentrations may have had a
greater effect, but it seemed implausible that dying cells would exhibit
metastatic behavior. In this regard, HGF blocked the effects of
dasatinib on viability at low drug levels, suggesting that the HGF
pathway modulates both viability/proliferation and invasion in this
model system. However, the absence of an increase in cell death or
apoptosis in response to 20 nM dasatinib (Figure 4) suggests that
dasatinib is primarily antiproliferative at low drug levels. In the
current study, we attempted to delineate between metastatic and
proliferative behaviors in vitro; however, the fact that TKI incubation
affects multiple physiologic pathways is important, and further
studies that seek to elucidate these differences are warranted.
HGF increased the rate of migration and invasion in both cell lines
tested. Similarly, both cell lines appeared to have constitutive
activation of the MET RTK at tyrosines 1230, 1234, and 1235,
which are in an activating kinase domain. Incubation with
low-concentration crizotinib, which is reported to be a MET-specific
TKI, or dasatinib did not significantly reduce the apparent
phosphorylation of MET at these sites. Future studies might benefit
238
Dasatinib Migration and Invasion in Canine Osteosarcoma
Marley et al.
by determining the extent and cause of the autophosphorylation of
MET in OS.
We were surprised that crizotinib but not dasatinib reduced the
HGF-induced secretion of active MMP-9 in the COS cell line.
Furthermore, there was no MMP-9 secretion detected from the
Clone-4 cells; yet the cells were presumably able to degrade the
invasion chamber substrate at the same rate as COS. We did not
directly assess the mechanism of protease activity in the invasion
model system but speculate that the Clone-4 cells depend on MMP-2
activity whereas the COS cells may achieve the same result using
MMP-9. Regardless, the present data show for the first time that OS
cells are able to secrete different forms of protease and these may
function to degrade extracellular matrix and facilitate invasion.
Our inability to show a difference in MET phosphorylation upon
activation with HGF suggests that HGF activated either a different
RTK or a different epitope of MET. HGF exists as an alpha- and
beta-chain heterodimer that can be cleaved into its component parts.
Ohnishi et al. [19] recently identified a new receptor for the
beta-chain of HGF (HGF-β), but this receptor was not investigated in
the current study. Further study is warranted to advance our
understanding of the mechanisms of HGF activation in OS.
The reported median survival time for dogs with appendicular OS
treated with amputation and carboplatin is approximately 10 months
[20,21]. Although the in vivo data presented here are anecdotal, we
feel that they help to inform the transition from in vitro investigations
of dasatinib in canine OS to the clinical setting. We previously
reported that 0.75 mg/(kg day) of dasatinib was tolerable and
achieved clinically relevant blood levels of the drug [16]. We have
now extended this to include three additional dogs that appear to have
improved survival. The one surviving dog reported here has lived
more than 3.2 times longer than the reported median survival time.
This dog has lived with waxing and waning pulmonary metastases for
most of this time. Seemingly, dasatinib has played a role in this
longevity. Although originally developed as a Src inhibitor, dasatinib
is now known to affect multiple kinases, each with a broad range of
functions. In this regard, dasatinib has been investigated for its
capacity to reduce the permissive effects of T-regulatory immune cells
in the tumor microenvironment [22–24]. We speculate that this
could be important to canine OS and feel that the current
observations, coupled with an acceptable dose and schedule for
dasatinib administration, provide a basis for a clinical trial in dogs
with OS.
Conflicting Interests
The authors declare no competing interests.
Acknowledgements
The authors thank the Canine Health Foundation (grant 01726A,
SCH) and the Oregon State University College of Veterinary Medicine
for their generous support of this research. The authors also thank
Dr Shay Bracha who participated in the care of patients in this report.
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
References
[1] Andermarcher E, Surani MA, and Gherardi E (1996). Co-expression of the
HGF/SF and c-met genes during early mouse embryogenesis precedes reciprocal
expression in adjacent tissues during organogenesis. Dev Genet 18(3), 254–266.
[2] Schmidt L, Duh FM, Chen F, Kishida T, Glenn G, Choyke P, Scherer SW,
Zhuang Z, Lubensky I, and Dean M, et al (1997). Germline and somatic
[23]
[24]
Translational Oncology Vol. 8, No. 4, 2015
mutations in the tyrosine kinase domain of the MET proto-oncogene in papillary
renal carcinomas. Nat Genet 16(1), 68–73.
Rong S, Segal S, Anver M, Resau JH, and Vande Woude GF (1994). Invasiveness
and metastasis of NIH 3T3 cells induced by Met-hepatocyte growth factor/scatter
factor autocrine stimulation. Proc Natl Acad Sci U S A 91(11), 4731–4735.
Gherardi E, Birchmeier W, Birchmeier C, and Vande Woude G (2012).
Targeting MET in cancer: rationale and progress. Nat Rev Cancer 12(2), 89–103.
Tseng JR, Kang KW, Dandekar M, Yaghoubi S, Lee JH, Christensen JG, Muir S,
Vincent PW, Michaud NR, and Gambhir SS (2008). Preclinical efficacy of the
c-Met inhibitor CE-355621 in a U87 MG mouse xenograft model evaluated by
18F-FDG small-animal PET. J Nucl Med 49(1), 129–134.
Michieli P, Mazzone M, Basilico C, Cavassa S, Sottile A, Naldini L, and
Comoglio PM (2004). Targeting the tumor and its microenvironment by a
dual-function decoy Met receptor. Cancer Cell 6(1), 61–73.
Logan TF (2013). Foretinib (XL880): c-MET inhibitor with activity in papillary
renal cell cancer. Curr Oncol Rep 15(2), 83–90.
Grullich C (2014). Cabozantinib: a MET, RET, and VEGFR2 tyrosine kinase
inhibitor. Recent Results Cancer Res 201, 207–214.
Stellrecht CM and Gandhi V (2009). MET receptor tyrosine kinase as a
therapeutic anticancer target. Cancer Lett 280(1), 1–14.
Sampson ER, Martin BA, Morris AE, Xie C, Schwarz EM, O'Keefe RJ, and
Rosier RN (2011). The orally bioavailable met inhibitor PF-2341066 inhibits
osteosarcoma growth and osteolysis/matrix production in a xenograft model. J
Bone Miner Res 26(6), 1283–1294.
McCleese JK, Bear MD, Kulp SK, Mazcko C, Khanna C, and London CA (2013).
Met interacts with EGFR and Ron in canine osteosarcoma. Vet Comp Oncol 11(2),
124–139.
Fieten H, Spee B, Ijzer J, Kik MJ, Penning LC, and Kirpensteijn J (2009). Expression
of hepatocyte growth factor and the proto-oncogenic receptor c-Met in canine
osteosarcoma. Vet Pathol 46(5), 869–877.
De Maria R, Miretti S, Iussich S, Olivero M, Morello E, Bertotti A, Christensen JG,
Biolatti B, Levine RA, and Buracco P, et al (2009). met oncogene activation qualifies
spontaneous canine osteosarcoma as a suitable pre-clinical model of human
osteosarcoma. J Pathol 218(3), 399–408.
McCawley LJ, O'Brien P, and Hudson LG (1998). Epidermal growth factor
(EGF)– and scatter factor/hepatocyte growth factor (SF/HGF)–mediated
keratinocyte migration is coincident with induction of matrix metalloproteinase
(MMP)-9. J Cell Physiol 176(2), 255–265.
Wang H and Keiser JA (2000). Hepatocyte growth factor enhances MMP activity in
human endothelial cells. Biochem Biophys Res Commun 272(3), 900–905.
Davis LE, Hofmann NE, Li G, Huang ET, Loriaux MM, Bracha S, Helfand SC,
Mata JE, Marley K, and Mansoor A, et al (2013). A case study of personalized
therapy for osteosarcoma. Pediatr Blood Cancer 60(8), 1313–1319.
Shor AC, Keschman EA, Lee FY, Muro-Cacho C, Letson GD, Trent JC,
Pledger WJ, and Jove R (2007). Dasatinib inhibits migration and invasion in
diverse human sarcoma cell lines and induces apoptosis in bone sarcoma cells
dependent on SRC kinase for survival. Cancer Res 67(6), 2800–2808.
Marley K, Helfand SC, Edris WA, Mata JE, Gitelman AI, Medlock J, and Seguin B
(2013). The effects of taurolidine alone and in combination with doxorubicin or
carboplatin in canine osteosarcoma in vitro. BMC Vet Res 9, 15.
Ohnishi H, Oka K, Mizuno S, and Nakamura T (2012). Identification of mannose receptor
as receptor for hepatocyte growth factor beta-chain: novel ligand-receptor pathway
for enhancing macrophage phagocytosis. J Biol Chem 287(16), 13371–13381.
Phillips B, Powers BE, Dernell WS, Straw RC, Khanna C, Hogge GS, and Vail DM
(2009). Use of single-agent carboplatin as adjuvant or neoadjuvant therapy in
conjunction with amputation for appendicular osteosarcoma in dogs. J Am Anim
Hosp Assoc 45(1), 33–38.
Bergman PJ, MacEwen EG, Kurzman ID, Henry CJ, Hammer AS, Knapp DW,
Hale A, Kruth SA, Klein MK, and Klausner J, et al (1996). Amputation and
carboplatin for treatment of dogs with osteosarcoma: 48 cases (1991 to 1993). J
Vet Intern Med 10(2), 76–81.
Fei F, Yu Y, Schmitt A, Rojewski MT, Chen B, Gotz M, Dohner H, Bunjes D,
and Schmitt M (2009). Dasatinib inhibits the proliferation and function of
CD4+CD25+ regulatory T cells. Br J Haematol 144(2), 195–205.
Yang Y, Liu C, Peng W, Lizee G, Overwijk WW, Liu Y, Woodman SE, and Hwu P
(2012). Antitumor T-cell responses contribute to the effects of dasatinib on c-KIT mutant
murine mastocytoma and are potentiated by anti-OX40. Blood 120(23), 4533–4543.
Kreutzman A, Ilander M, Porkka K, Vakkila J, and Mustjoki S (2014).
Dasatinib promotes Th1-type responses in granzyme B expressing T-cells.
Oncoimmunology 3, e28925.